In this paper, a new approach to Mediterranean Water Eddy border detection is proposed. Kohonen self-organizing maps (SOM) are used as data mining tools to cluster image pixels thr...
It is often reported in the technique literature that the success of the self-organizing feature map (SOM) formation is critically dependent on the initial weights and the selectio...
Abstract. The use of self-organizing maps to analyze data often depends on finding effective methods to visualize the SOM's structure. In this paper we propose a new way to pe...
Self-Organizing Maps (SOM) are very powerful tools for data mining, in particular for visualizing the distribution of the data in very highdimensional data sets. Moreover, the 2D m...
— Many information retrieval and machine learning methods have not evolved in order to be applied to the Web. Two main problems in applying some machine learning techniques for W...